Structure and agency
The starting point of this topic is that while a lot of research has been concentrated on particular aspects or scales of the kinds sketched so far, rather little has been offered by way of integration. It is an appropriate time to try, as these sorts of issue have begun to be widely debated. Ian Hodder has drawn attention to the need in studies of the British Neolithic to take account of individual action (1999, 132—7; cf. Hodder 1986), and Mark Edmonds has written an unusually empathetic account of the southern British Neolithic (1999). The relationship between what have come to be called structure and agency has begun to be widely debated in archaeology (e.g. Hodder 1999; Dobres and Robb 2000a; cf. Layton 2000), typically some time after the initial debates in the field of social theory and social anthropology. The principal sources have been Bourdieu and Giddens, each in turn influenced by a number of earlier and contemporary writers, including Mauss in the case of Bourdieu, and Bourdieu, Goffman and Garfinkel among many others in the case of Giddens.
Bourdieu was concerned with daily action and the engagement of the body in it, with the minutiae of life as well as the realms of more conscious thought (Gosden 1999, 125). This was expressed in his important concept of habitus, the bare definition of which is somewhat forbidding at first sight: systems of durable, transposable dispositions, structured structures predisposed to function as structuring structures, that is, as principles of the generation and structuring of practices and representations which can be objectively ‘regulated’ and ‘regular’ without in any way being the product of obedience to rules, objectively adapted to their goals without presupposing a conscious aiming at ends or an express mastery of the operations necessary to attain them and, being all this, collectively orchestrated without being the product of the orchestrating action of a conductor. (Bourdieu 1977, 72)
Habitus is a generative or enabling disposition, generally absorbed or learnt unconsciously; it is a second nature and a ‘feel for the game’; ‘people produce thought, perception and action without thinking about how they are doing so, but in a manner which has its own logic’ (Gosden 1999, 125-6). As Bourdieu also puts it (1977, 78), habitus is ‘the durably installed generative principle of regulated improvisations’. Agents or subjects, both terms used by Bourdieu, are in part portrayed as not really knowing what they are doing:
It is because subjects do not, strictly speaking, know what they are doing that what they do has more meaning than they know. The habitus is the universalizing mediation which causes an individual agent’s practices, without either explicit reason or signifying intent, to be none the less ‘sensible’ and ‘reasonable’. (Bourdieu 1977, 79)
Habitus, finally, is historically grounded; ‘the system of dispositions . . . is the principle of the continuity and regularity which objectivism discerns in the social world without being able to give them a rational basis’ (Bourdieu 1977, 82).
Bourdieu was concerned to go beyond the opposition between ‘subjectivism’, or the search for how people experience and conceptualise the world, and ‘objectivism’, the search for underlying structures, independent of people’s knowledge, concepts or purposes (Baert 1998, 30). In a sense, this was also the goal of Giddens (1979; 1984), but is expressed by him in a wider theory of action, which draws not only on Bourdieu but also on a number of other sources, and is presented as applicable to a broad range of historical situations, especially recent ones in complex modern societies, as opposed to the more circumscribed settings analysed by Bourdieu. For Giddens, agency is a blend of tacit knowledge and practical consciousness (Baert 1998, 105); agency is not necessarily the same as intentionality, ‘purposive’ rather than ‘purposeful’ (Baert 1998, 101; Giddens 1984, 9). Agency is a ‘continuous flow of conduct’ (Baert 1998, 101), ‘a stream of actual or contemplated causal interventions of corporeal beings in the ongoing process of events-in-the-world’ (Giddens 1979, 55). There is also the conviction that people are knowledgeable and are always able to act otherwise (cf. Baert 1998, 101). Structure is marked out from system (the patterning of social relationships across time and space), and defined as ‘a set of social rules and resources which are recursively implicated in interaction’, and ‘located outside space and time, existing only in a virtual way as memory traces, to be implemented in spatially and temporally located interaction’ (Baert 1998, 101—2). At the heart of this model is the idea of the duality of structure and its realisation in structuration (Giddens 1984, 25). ‘Structure allows for agency, which in turn makes for the unintended reproduction of the very same structures’; ‘structures, as rules and resources, are both the precondition and the unintended outcome of people’s agency’ (Baert 1998, 104).
This range of ideas has been widely influential, but should not be treated uncritically, especially since there is a tendency for archaeology to pick up theories from other disciplines some time after their original formulation and discussion. Both Bourdieu and Giddens are better at describing some of the conditions in which social order is reproduced than at dealing with change (cf. Baert 1998, 33 and 110). Both downplay other, more conscious forms of thought, whether or not these characterise or are part of the routines of daily life. Despite drawing in part on Bourdieu, Giddens presents rather disembodied agents. Both present a rather shadowy picture of structure or structures, and the distinction in the case of Giddens between structure and system is artificial and reductionist.
The number of archaeological studies taking this body of theory as a starting point has multiplied recently. Barrett (1994) discussed the notion of knowledgeable agents in an extended case study of the Avebury region, but these agents remain somewhat faceless, lacking motive or value. Dobres (2000) has attempted to bring people back into the study of technology, within the context of a wide discussion of agency. There is less direct discussion of structure or context, and the detailed case study here (Dobres 2000, 187—209) consists essentially of a demonstration that in a specific late Palaeolithic regional context there was variability in artefact manufacture and repair, which may relate to age and gender differences on the one hand, and perhaps to a non-aggressive use of material culture as a means of social display on the other. One of the more detailed discussions so far has been of the conditions in which changes in container technology were adopted in one region of the eastern United States, from about 4500-3000 BP (Sassaman 2000). Starting from the proposition that ‘actions to build consensus or norms are likewise agential in that they derive ultimately from efforts to create rules or traditions in opposition to other structures’, Sassaman (2000, 149) argues that there was considerable conflict between insiders and outsiders, and that women were of central importance in a residential system with matrilocality and unilineal descent. Women may have made pottery, while men were committed to an older tradition of soapstone cooking slabs. In time, soapstone vessels appeared, possibly to assert more inclusive relations among regional populations (Sassaman 2000, 163). Agency can be seen in collective action and in collectivities. It is not clear in this account whether agency is to be seen as consistently conscious and intentional, or unconscious and habitual. There seems no reason why it should not have been both, and a very general term is being called upon to cover a very wide range of situations (rather as though we were to approach the study of literature with a central concept of literacy).
While attention to the structure-agency debate is undoubtedly proving fruitful in archaeology, these examples also illustrate some of the difficulties as far as archaeology is concerned. Perhaps at least three recurrent problems can already be identified. First, there is a tendency to brevity. Sometimes this takes the form of programmatic statements, with little or nothing by way of supporting case studies. Thus Hodder has used the convenient example of the Ice Man to illustrate his view of how individual agents could be seen to fit into and influence the historical situations to which they belonged (1999; 2000). The Ice Man is probably in fact one of the least good examples to choose, despite the immediacy of his presence as given by the remarkable circumstances of preservation, since his death was isolated and it is hard to relate him to his contemporary context. Besides, it is equally legitimate to see him as a routine figure, of a kind long familiar in high places, rather than as a new kind of persona symbolising the emergence of a more aggressive, separate and male-oriented world.2 In either case, the argument needs to be much longer. Another, thought-provoking, programmatic statement on agency distinguishes between structural conditions and structuring principles, to emphasise the distinction between daily routines and what people thought and valued, but no examples are given (Barrett 2000, 65).
Secondly, there has been a tendency merely to substitute one scale of analysis for another: simply shifting from one end of the telescope to the other. This is not confined to archaeology. Marilyn Strathern has commented on the strong tendency in social anthropology to offer ‘single relationships’ at the expense of more complex, overlapping and sometimes contradictory ones (1987, 29), and J.F. Weiner has fundamentally criticised the ‘belief that cultures, particularly those of small-scale societies, can be typified by a single set of structured propositions’ (1988, 1). Archaeologists have been prone to substitution: in differing ways over recent years for example, meaning for function, symbolism for practicality, mobility for seden-tism, and so on. In a move away from general process and the operation of systems, the individual has been emphasised and individual voices sought (Hodder 1986; Tringham 1991). The gender of whole sets of individuals has been sought, and even where the concern for gender alone has been extended to a wider interest in ‘life-process’ (Derevenski 2000), analysis has tended to seek the signature of the individual life course, rather than simultaneously embrace the history of the collectivity to which individuals belonged.
Thirdly, the very terms of the structure-agency debate tend to favour abstraction. Both structure and agency tend to be conceived in abstract ways, whether in the distinction already mentioned between ‘structural conditions’ and ‘structuring principles’ (Barrett 2000, 65) or in the definition of habitus already cited. Agency is often unhelpfully seen as something universal (Gero 2000, 37), and collective agency is often ignored (Sassaman 2000, 149). Agency is often seen in opposition to structure (Moore 2000, 260), and quite clearly agency can be and has been defined in multiple ways (Dobres and Robb 2000b). Such a multiplicity of valid definitions for a single very general term in itself indicates the difficulties. If ‘culture’ should not be seen as something monolithic (Kuper 1999), then a far wider range of agencies, as the conditions and capabilities for action, must also be considered.
Some of the initial outcomes of this debate, at least as far as prehistoric archaeology has been concerned, could be summarised briefly as follows. People, whether as individuals or collectives, act within settings, practices and situations. Some of their actions are conscious and deliberate, others routine and almost unthinking. Some are strongly guided if not determined by the context, whereas others may be aimed at changing, challenging or otherwise subverting the context. Sometimes people act on their own, but it is clear that agency is not to be reduced only to individuals; in other cases, groups or collectives of people may act together. The effect of action depends on context. Where the setting is stable and values strongly normative, individual or other agency may serve mainly to maintain what was already in place, where less stable and with values and goals fluid, it may lead to rapid change. There may also be unintended consequences to action. In no situation is it really legitimate to separate agency and structure, since one does not exist without the other.
It is not that this debate has been unhelpful, but rather that in archaeology it has generally not gone far enough beyond substitution and abstraction, to allow better demonstration in particular cases of these general points. Because the debate has so far largely failed to engage in detailed studies, it has also failed to give any sort of nuanced sense of what it was like to be an acting person or agent in times remote from the present. A wider discussion is needed, if a better sense of being there is to be achieved. There is great scope for a closer look at the unglamorous routine of many aspects of past lives. This will bring in a discussion of what has been called the taskscape and the dwelling perspective, a sense of people bodily engaged in a world that was never separate or pre-formed, awaiting their interpretation of it. Ingold, for example (1996, 120-1), has suggested that ‘apprehending the world is not a matter of construction but of engagement, not of building but of dwelling, not of making a view of the world but of taking up a view in it’. On the other hand, this kind of treatment of the habitus can ignore the worldview of people. How people take a view of the world may depend crucially on how they are encultured. It is dangerous to ascribe too much to culture (Kuper 1999), but perhaps equally unwise to leave it out of account altogether (Sahlins 1999). As one particular dimension of this, values, ideas, ideals and emotions could be seen as the framework of a moral network or a moral community, and constitute a vital part of the ‘structures’ within which agents act, though one largely ignored so far by prehistoric archaeologists. (One exception, perhaps, is the term ‘structuring principles’ advocated by Barrett (2000, 65), though only in a general discussion.) Thirdly, all these possibilities need to be illuminated by a much better sense of how people thought. Debate has so far mainly concentrated on a rather simple distinction between conscious and unconscious thought, the latter often restricted to the operation of routine. ‘What goes without saying’, however, may encompass major issues and central concerns (Wagner 1967, 223; Bloch 1998). In any given situation, people may hold a range of overlapping and sometimes contradictory ideas, in what has been called ‘hybridisation’ (Latour 1993). There is a tension between the possibility that ritual, for example, is a formalisation or intensification of other activities (Bell 1992), and the likelihood that myth, as another example, exists as a separate domain, a valid but alternative ‘as if’ within particular settings (J.F. Weiner 1988). Many worldviews might be seen as a series of central concerns, rarely expressed as a whole and held in non-linear fashion (Bloch 1998). And finally, there is the opportunity to give a more specific sense to both ‘agency’ and ‘structure’ by considering what constitutes individualism and identity on the one hand, and networks of interaction (Latour 1993; M. Strathern 1996) on the other. There is much still to consider.
Taskscape and dwelling
As already noted, there has been rather general discussion (drawing on Bourdieu and beyond him Mauss, and other sources) of the habitus, the setting of existence in which habits and unthinking bodily action maintain a sense of understanding of the world.An overlapping recent strand has been the ‘dwelling perspective’, derived in large part ultimately from Merleau-Ponty (Tilley 1994, 12-14) and Heidegger (Thomas 1996), and further considered in a series of important papers by Ingold (1993; 1995; 1996; cf. Harris 1998; 2000). This, parallel to the idea of habitus, discusses how people act in a world which is never separate, pre-formed or a prior given. The dwelling perspective (Ingold 1993; 1995; 1996) seeks to give a better sense of how people get on in a world which is not pre-given, ‘taking the human condition to be that of a being enmeshed from the start, like other creatures, in an active, practical and perceptual engagement with constituents of the dwelt-in world’ (Ingold 1996, 120-21); according to it, ‘knowledge of the world is gained by moving about in it, exploring it, attending to it, ever alert to the signs by which it is revealed’ (Ingold 1996, 141). This kind of approach conveys well a sense of people attending to their world, in ways which may not have changed much over long periods of time. As Bloch has noted (1998, 5), we are sometimes guilty of seeking too much diversity, a question that will be one of the themes of this topic. Paradoxically, diversity may be a constituent of long-term stability.
The initial formulation of the taskscape ended with a scene taken from a Bruegel painting, The Harvesters. Ingold (1993, 164-71) uses this to show how people might have attended to a series of social relations while the task of getting the harvest in was being carried out, in a landscape formed by experience and movement through it. The challenge remains of using the approach with archaeological evidence. Edmonds (1999) has discussed the woodland setting, and in a general way the use of animals, but I can think of no detailed accounts of the routine socialities that may have been played out in either plant cultivation or animal husbandry. This then is one challenge which should be attempted with archaeological evidence.
The approach is also incomplete. It seems to give insufficient attention to learning and to socialisation. These may be long processes during childhood, in their turn almost unconscious or casual (e.g. Mead 1943), but there are also stages, such as initiation, when instruction may be much more direct, and when, as among the Hua of Eastern Highland New Guinea (Meigs 1990), gender ideologies may be at their most accentuated. The taskscape approach seems to give insufficient attention to the weight of collective tradition or culture (cf. Sahlins 1999) — however that may be taken up or contested by individuals — which may affect how people acquire ‘the skills for direct perceptual engagement’ (Ingold 1996, 142) with the world. It seems an extreme claim that people never make a view of the world, and it is possible to propose that people act at different times and in different situations from varied perspectives. The Harvesters painting itself may promote a certain moral stance, a particular view of the world. The dwelling perspective is best at giving a general sense of the flow of life, but less satisfactory at showing how people can do basic things in very different ways, or how they cope with innovations. The issues of diversity and uniformity, and of change and stability, recur again.
Culture, values and the moral network
Culture has been having a bad time in some circles, but lives happily on in others. Prehistorians of a theoretical bent have largely ignored it since the 1960s and 1970s, culture in some collective sense being replaced by material culture, active, almost an agent in its own right, particular, used to contest and subvert as much as to bind or to express either unthinking or conscious solidarity. This mirrors the diversity of use of symbols in cultural studies. A recent critique by a British social anthropologist of the American tradition of cultural anthropology recognises the general sense of culture as simply a way of talking about collective identities (Kuper 1999, 3), though it proposes that in the end the concept is an unhelpful (and unhealthy) lumping together of too many disparate processes (Kuper 1999, 247).
Such diversity and many-strandedness are one of the main themes of this topic, but it seems useful to revive interest in the ties that bind, and in a sense of collective agency. It has been strongly argued that even the ‘codgers’ of the American cultural anthropology tradition such as Boas or Linton did not see culture as ‘universally shared, monolithic or otherwise coherent socially or consistent logically’ (Sahlins 1999, 405). Wagner has described, with reference to the Daribi people of New Guinea, how, though alliances can take many forms on the ground, ‘community’ is a popular tactic of clan interaction and principles of exchange, expressed as belief in consanguinity (1967, 211, 216). But even if invented and tactical, a sense of collective identity can nonetheless be very real, especially in small-scale societies. We shall come back to questions of who might have been included in the varying uses of ‘we’: ‘We, the LBK people’ as well as ‘We, the Tikopia’?3
As one way of talking about collective identities, I should like to focus more attention on values, ideals and emotions. These are important, as they are all in a sense ideas, which are unlikely in any case to have been generated by single individuals, and the scale of collectivity involved is left open. Most of the individuals sketched so far by an interpretive, post-modernist archaeology lack any or much sense of shared values. While the domus can indeed stand as a quite rare example of an explicitly formulated value system that includes everyone involved, the same cannot be said of the agrios, which only seems to have affected adult men (Hodder 1990). The Ice Man’s self-sufficiency is enough to make him stand as a representative of a whole ideology (Hodder 1999). I have argued instead that there was a long-lived general set of values in the European Neolithic, incorporating ideals of participation, sharing, non-accumulation and commonality but also the pursuit of prowess.This is not to claim that values in the European Neolithic were uniform in all times and places, or that actual behaviour always accorded with such ideals. Nor is it to restate a Durkheimian position in which society makes its individuals. But there is socialisation, and each individual does not take up a view in the world de novo and unaffected by others; what others have done in the past and are doing in the present must affect how people acquire ‘the skills for direct perceptual engagement’ (Ingold 1996, 142) with the world, unless we are to argue that engagement with the physical and social world is something of universal character.
Analogies suggest a wide range of possibilities. In an Aristotelian sense of being concerned for the response of others, it is possible to see exchange for example as a moral activity (Hagen 1999). In arguing that exchange among the Maneo of eastern Indonesia (note the first of many examples where it seems to make perfectly good sense to refer to collective identity, even though this can also be broken down or analysed at other scales) has in this sense a vital moral dimension, not reducible to its social effects, Hagen has suggested that the Maneo are not guided by specific moral principles such as would mandate sharing (1999, 362). In many cases in the historical or recent Mediterranean world, individual personal responsibility is held in the ‘triangle of honor, shame and luck’ (Douglas and Isherwood 1996, 23). Referring to Kabyle society in Algeria, Bourdieu (1977, 48) has observed:
The ethic of honour is the self-interest ethic of social formations, groups or classes in whose patrimony symbolic capital figures prominently. Only total unawareness of the terrible and permanent loss which a slur on the honour of the women of the lineage can represent could lead one to see obedience to an ethical or juridical rule as the principle of the actions intended to prevent, conceal or make good the outrage.
The western terms ‘ought, ‘want’ and ‘must’ may not be easily applied to the Daribi of New Guinea, whose moral law is based on ‘relative, political intergroup situations’ (Wagner 1967, 29). In other cases, however, imagined moral life and moral language seem to be more centrally linked to an ethical sense, tied to ideas of the person and identity, and generating powerful emotions within a shared value system. Among the Amuesha people of central Peru, greediness and meanness are regarded as immoral, irrational and antisocial, and ‘power is legitimate only when its holders are seen as loving, compassionate and generous life-givers’ (Santos-Granero 1991, 229). Among the Rauto people in Melanesia, ceremonial exchange carried out in the right way creates a ‘cultural landscape of memory and emotion’, and the emotions generated can be considered as a kind of moral perception; emotions ‘define and render compelling a particular moral stance towards life’, the result of choice between this and more individualistic alternatives (Maschio 1998, 86, 97). Among the Western Apache, as described by Basso (1984), certain historical narratives served to link past events to named places, in stories which ‘stalked’ their listeners with their moral force; and native American appropriation of the landscape has been envisaged as being effected by ‘an act of imagination which is moral and kind’ (Momaday 1976, 80). These are all examples of values and emotions with a central place in social relations and living in the world. Among the Etoro of Papua New Guinea, however, male-dominated social inequality is constructed as a moral hierarchy or hierarchy of virtue, grounded in cosmology and worldview (Kelly 1993). A moral sense need not in itself be neutral, and there is certainly no need to idealise or whitewash it. Concerning the Daribi again, Wagner (1967, xxviii) has written of ‘their dour, almost boastful pride in a culture which is a curse, whose demons are named for diseases, whose powers are the ill-wills of the wronged, whose triumph is the negation of a negation’; the individual Daribi reacts with anger to the ways in which the outside world possesses the soul and may bring sickness, weakness and madness (Wagner 1967, 42).
The concept of a ‘moral community’ has been used in a discussion of the Nuer, to connote those participating in a common value system (Johnson 1994, 327—9), as well as with reference to the Amuesha (Santos-Granero 1991, 119); the related concept of mutuality has also been discussed by Moore (1988) and Gosden (1994), and the practice of ‘moral coalitions’, though with a greater sense of conflict between sets of gender-based values, has been discussed by Robb (1994). Adapting these ideas, the idea of a moral network can be proposed; the extent to which others are involved and affected may define the moral network, which like other networks is liable to remain local at all points (Latour 1993, 117—20) but may also be open and unbounded. Values may be seen to act as sanctions on behaviour; there were limits to what individuals or limited interest groups could attempt or hope to get away with, and there may have been limits, within this perspective, to what they could conceive as possible. In this way, we can perhaps begin to put some flesh on what Barrett has called ‘structuring principles’ (2000, 65).
How they thought
It is not enough just to suggest kinds of thought and value, as it were alongside patterns of bodily action. It is also necessary to pay more attention to how those were expressed. This may vary. There may be at least four significant aspects: non-explicit performance of ideas, non-linear expression based around central nodes or concepts, alternative ‘tropes’, and historical hybridisation. The dwelling perspective discussed above evokes a rather active, conscious kind of attention to the world. This may not be how people think all the time. The anthropologist Maurice Bloch (1992) has explored a sense of ‘what goes without saying’ in Zafimaniry society in Madagascar, by looking at attitudes and beliefs, rooted in practice and material experience, which are central to people but which seem so obvious that explanation of them to outsiders seems pointless. These concern ideas about such subjects as people themselves, trees, sex, gender and houses. The relevance of elaborate house decoration is not easily put into words – what goes without saying – as it is simply part of the right way to treat the living and growing house (Bloch 1995a). Once again, some kind of collective dimension is involved. Not all relevant ideas can be seen as confined to the practice and material experience of a single generation. The concept of the maturing house (in which the hardening of its wood is a metaphor for the growth and success of the household) relies on a considerable passage of time and presumably some active transmission from generation to generation. However it is passed on, this is also a powerful model for thinking about enduring beliefs over very long periods of time (even though the historical Malagasy situation has been far from static or timeless: Bloch 1998).
Another dimension of this study was a use of ‘connectionism’ (see also Bloch 1993), or the non-linear ways in which people often actually seem to think. Speed of thought and the ability to react instantly in different social situations suggest that thought can be held in central nodes or concepts, such as again in the Zafimaniry case to do with what people are like and how they mature, the differences and similarities between men and women, and what good marriages, trees and wood, and houses are like (Bloch 1992). In the Zafimaniry case, these could be seen to make up some sort of coherent worldview, though that might never or rarely be expressed as a single unified whole. Or at least, there is little obvious contradiction or opposition between these suggested nodes of thought.
In other cases, the propositions of a ‘culture’ may exist on different planes, as it were on separate drives as well as in different folders. This has been explored with reference to myth, song and dream by J.F. Weiner among the Foi people of Papua New Guinea (1988; 1991).Weiner argues for a number of views of the world, a ‘series of successively embedded metaphors’ in which central concerns with male-female relations can be played out (1988, 285). Myth does not serve, however, to mirror exactly the facts of daily social existence, but is another and equally valid rendering of those (J.F. Weiner 1988, 5, 16). In a slightly different way, song and dreams open up different ways of looking at and re-ordering the world. For the Foi, it is important to communicate with the ghosts of the dead in dreams, an activity in which males are often engaged (J.F. Weiner 1991, 1-8). These ghosts, however, are ambiguous and elusive, inhabiting an upside-down world, part ill-disposed towards the living, part benevolent. Men seek a link with the power of ghosts through dreams. Women also have a central role in looking after the newly dead. It is they who create songs, which link names and places in the flow and movement of Foi life (of which more later), but it is men who perform such memorial litanies, preferably in the intense and public sociality of their longhouse villages, rather than in more isolated and ghost-haunted bush houses. Knowing the world is thus a complex process, involving several concepts or nodes, the dead as well as the living, play between genders, and gendered creativity and performance.
Compartmentalisation of thought may also in part be the outcome of histories. In a discussion of post-1989 Europe, it has been proposed that we do often think in compartmentalised ways, a pattern or mixture which has been called hybridisation (Latour 1993). At the present time, we may be said to retain elements of pre-modern, modern and post-modern thought (Latour 1993, fig. 5.1). This model certainly fits the discipline of archaeology in its present state. Thus the discipline still happily encompasses many concepts based on evolution (from the nineteenth century), regularly uses the concept of process from the modernist phase of the 1960s and 1970s, and has engaged since the early 1980s in post-modernist self-reflexivity; only the manifesto writers see themselves as outside such a mixture of approaches. This is a very general but potentially important and useful way to think about modes of thought in the past across horizons of change and over long periods of time. Bloch has noted one small example of this, in the way in which some Malagasy farmers speak to their animals not in their native tongue, but in French, the language of colonial power (1998, 193—5).
Individuals and networks
We keep coming back to questions of scale. The discussion of the taskscape in the dwelling perspective, or of the habitus and its nexus, the body, offers the possibility of looking closely at individual action. Many of the basic routines of past existence may have been persistent, and widely shared. They may also have varied subtly from situation to situation, encompassing both diversity and the potential for change. The tendency so far has been to treat agency as individuality, in the terms of Rapport (1996), or the ability to act, rather than as individualism, or the potential to act in distinctive ways.There has been little attempt to consider the very varied ways in which identity and concepts of the individual may have been constituted, and their significance for questions of diversity and change.
On the other hand, ‘structure’, in the varied guises of ‘society’, ‘community’ and ‘archaeological culture’, has generally been seen as a closed system, and the challenge has been conceived as one of working out the play between internal dynamics and external pressures. In this regard, the alternative notion of network may be more helpful, as it leaves the question of closure unanswered. The notion of closed system has been a powerful one, including in social anthropology. With reference to descent, it has been shown how the model of closed, unilineal descent was for a very long time forced on to the ethnographic evidence, suiting our own view of the world better than the evidence on the ground (Kuper 1988). The idea of network raises other possibilities, with open – perhaps bilateral – systems of descent, ties that go on for ever, and alliance and co-residence at least as important as kinship and descent. Leach argued such a position for the Kachin of highland Burma (E. Leach 1954), suggesting, as Tambiah (1998, 312-13) has summarised it, ‘an open system of many lineages linked in circles of wife givers and wife takers, communicating with one another diacritically through variations of dialect, dress, and other local differences, and capable of dynamically generating . . . as well as contesting tendencies towards extra-local hierarchical political formations’. But it has taken a long time for archaeology to consider its implications (see also E. Leach 1973). It has been argued that even the longest networks (from global corporations to railway systems) can be seen as local at particular times and places (Latour 1993, 117), and that we are accustomed to cutting the network, in order to make sense of things within the framework of our own lives (M. Strathern 1996). These abstract propositions should be regarded with suspicion. We cannot generalise the character of every network in advance. Whether situations were seen as purely local, or whether the local was strongly bound into factors much further afield, requires careful examination in specific instances. Much may depend on the stage in the ‘life course’ (cf. Derevenski 2000) of individuals concerned and the kinds of sociality possible or required.
The Ice Man provides an initial, limited example. It has been argued that his way of life increasingly stood for something outside the bounds of lowland agricultural life in the later fourth millennium BC, his independence and self-sufficiency representing the separate male world of the agrios (Hodder 1999; 2000). Seen as part of a network or perhaps better a series of networks, the situation may seem rather different. Many people before him had gone into and across high places. TransAlpine contacts go back to the early Holocene period (though they probably did gradually intensify through time). There is some general support for the notion that males among foragers can range quite widely, at a lifetime scale, in search of mating partners.The Ice Man was clearly in contact with lowland communities. He must therefore have been enmeshed in various networks, and habituated to different kinds of sociality in varying settings, places and seasons. Perhaps such skills came best to an older person, with more experience. Whatever precisely he was doing on the trip that led to his death, his existence cannot be reduced to a single dimension.
The argument that follows
These preliminary concerns define the argument that follows.Routines are never neutral; they are embodied ways of dealing with the world. They thus show how much was held in common in daily life, but also the possibility for diversity, not least because acting in the world, being there, requires the presence of others and demands sociality. Paradoxically, from this structural condition may have come not only widespread diversity but also long-term stability.This also involves consideration of values and the moral community, and I shall argue that this was another factor in long-term stability.These are mainly from early situations, and will bring us back, among others, to one of the examples with which this topic started. In the last brief topic, I take a longer time perspective, and suggest that the many-sidedness of past existence is part of why things may have changed more slowly than we have been accustomed to think.